I still remember the first time a private-banker friend described a product that could hand a client a six-figure ordinary deduction in year one while the account itself stayed invested. He lowered his voice the way people do when a trick feels a little too clean. Three years later that whisper has turned into a formal warning from the tax authorities, and the part that should keep family offices awake is not the theory. It is the line that says future guidance could reach back and touch trades already done.
For a stretch, the hottest allocation on parts of Wall Street was not a chip stock or a private-credit sleeve. It was a carefully built pattern of losses. The portfolio was meant to keep compounding. The tax character of those losses was meant to land where it hurt the Treasury most: against salaries, bonuses, and other ordinary income taxed at the top rates. On Monday, Treasury and the revenue service answered with a paired notice and ruling. Call it a warning shot, not a final verdict. Comments run through late October. The retroactivity language is already doing the real work.
Why Engineered Ordinary Losses Suddenly Look Fragile
The notice reads like a catalog of structures the market had treated as clever rather than controversial. Officials say they are studying, and may later designate as listed transactions or transactions of interest, a cluster of techniques that manufacture a mismatch. Gains show up as capital. Losses show up as ordinary. That split is the whole game. A capital loss mostly helps someone who already has capital gains. An ordinary loss can shelter a paycheck.
Perhaps the most interesting aspect is how openly the pitch had been framed. This was not hidden in a footnote. It was the product. A client writing a large check could, on the numbers some funds reported, book ordinary losses equal to a big slice of capital in the first year, then stay in the strategy so the winners kept running. One well-known tax-aware long-short vehicle, with about $6.6 billion at midyear, recorded ordinary losses in 2025 equal to roughly 28 percent of capital invested. Put a $10 million subscription next to that ratio and you get something near $2.8 million of losses to set against wages. The account was designed not to look like a busted trade. It was designed to look like a deduction factory that still made money.
I have found that investors hear “loss” and assume pain. Here the loss was often an accounting result of how positions were opened, closed, or characterized, not a collapse in net worth. That gap between economic result and tax result is exactly what the authorities say they are examining.
What The Notice Actually Puts On The Table
The text is broader than one fund. Treasury is looking at tax-aware funds that pair capital gain with ordinary loss. It is looking at same-day buys and sells of foreign-currency forwards, the sort of trade that can drag Section 988 character into the conversation. It is looking at selective terminations of notional principal contracts, which in plain English means equity swaps closed on a timetable that suits the tax year more than the investment thesis. Identified straddles with mixed character are in the pile too, the swaps-plus-futures combinations that reporters had already flagged.
Box-spread exchange-traded products that mimic Treasury-bill returns without throwing off current income are mentioned. So are assorted games around the regulated-investment-company qualifying-income test, including the use of in-kind redemptions. None of this is a final ban. A notice is a study, a request for comment, and a signal. The signal is loud.
- Capital-gain and ordinary-loss mismatches inside tax-aware funds
- Same-day foreign-currency forward round trips
- Selective close-outs of equity swaps and mixed-character straddles
- Box-spread products that resemble cash yields without current income
- In-kind redemption patterns that strain fund qualification rules
Officials did leave a door open. Plain long-short stock strategies, they conceded, may fit long-standing techniques. The objection is aimed at the ordinary-loss machinery, and at funds that look primarily tax-motivated rather than aimed at an economic return from genuine investment activity. That sentence is doing a lot of work. It is also, frankly, close to the marketing line some of these products used.
The Retroactivity Line Is The One That Matters
Guidance that “could apply retroactively” is not the same as guidance that will. Lawyers will parse that word for months. Still, by refusing to grandfather existing trades, the authorities have put a question mark beside every recent partnership schedule that shows outsized ordinary losses. A K-1 is not a rumor. It is a number a client already used, or plans to use, on a return.
In my experience, sophisticated investors tolerate complexity when the downside is a lower return. They tolerate it less when the downside is an amended return, interest, and a fight over penalties. The notice does not impose those outcomes today. It makes them imaginable. That is enough to change behavior.
When a deduction depends on character the statute did not obviously invite, the cleanest risk is not the market. It is the possibility that the character gets rewritten after the year has closed.
Seasoned tax counsel, paraphrased from client briefings this week
Comments are due by October 28. That window is short for a structure that took years to industrialize. Firms will file letters arguing economic substance, investor intent, and the difference between a hedge and a harvest. Some of those letters will be serious. Some will be stalling. The calendar does not care which.
How A Niche Became A Balance-Sheet Item
Step back a year and the scale stops looking like a boutique trick. More than a trillion dollars has been described as sitting in tax-efficiency strategies of one kind or another, from direct indexing at the mild end to leveraged long-short books at the sharp end. Direct indexing is familiar. You own the stocks, you harvest losses when names dip, you try not to drift too far from the index. The exotic cousin is different. It uses leverage, shorts, and sometimes derivatives so the loss stream can be larger, faster, and pointed at ordinary income.
One large quantitative firm rode that demand hard. After a rough stretch that left assets under $100 billion a few years ago, tax-aware products helped pull the firm back toward the top of the hedge-fund tables. Long-short tax assets were described as moving from about $3 billion in 2023 to roughly $70 billion, around 40 percent of the firm. That is not a side sleeve. That is the business.
An aggressive variant, according to figures circulated in market write-ups, could turn a $100 million investment into more than $580 million of tax-offsetting losses over a decade. Read that twice. Losses nearly six times principal, while the account is still up. You can argue the math. You cannot argue that the pitch was subtle. A short seller who had been betting authorities would step in called ordinary-income shielding the industry’s holy grail. A tax professor was blunter still: if nothing is done, the addressable market is enormous, closer to carried interest on steroids than to a tidy planning idea.
I do not love the steroid metaphor, but the logic is hard to dodge. Carried interest is a fight about how a slice of fund profit is taxed. Ordinary-income sheltering is a fight about the wage base itself. Surgeons, law-firm partners, and senior bankers do not all have huge embedded stock gains. They all have ordinary income. That is a much wider door, and a much larger hole if the door stays open.
Capital Gains Harvesting Versus Wage Shelter
It helps to separate two ideas that marketing often glues together under the label tax alpha. Classic loss harvesting mostly rearranges capital gains. You sell the loser, keep the winner, and try to avoid a wash sale. The benefit shows up if you have gains to offset, or if you can bank a limited ordinary deduction and carry the rest. Useful. Bounded. Old.
The newer pitch aims higher. It wants losses that are ordinary, so they can offset wages and bonuses taxed at the highest marginal rates, while gains that do arise stay capital and wait for a lower rate or a later year. That is not the same product with a better brochure. It is a different claim on the tax code.
| Approach | What It Usually Offsets | Who Feels The Benefit | Policy Heat |
| Direct indexing | Capital gains, limited ordinary slice | Investors with appreciated portfolios | Lower, long accepted |
| Stock long-short harvesting | Mostly capital losses | Taxable accounts with embedded gains | Mixed, often tolerated |
| Ordinary-loss overlays | Wages, bonuses, other ordinary income | High earners even without big gains | High, now under study |
| In-kind fund conversions | Built-in gain on contributed stock | Holders of concentrated winners | High, ruling already out |
The table is a simplification. Real funds blend these rows. Still, the policy temperature rises as you move down. Capital-gains management mostly helps people who already won in the market. Wage sheltering helps people who are simply well paid. An administration funding a large deficit, with the ten-year yield having briefly topped 5 percent for the first time since 2023, can do that arithmetic without a seminar.
Warning Lights Were Already On
This did not arrive from a clear sky. Treasury officials said at a New York conference in July that some structures produce outcomes Congress did not intend and were potentially abusive. That is conference language, not a statute. Markets heard it anyway. Two large retail brokerages quietly started limiting new accounts chasing the strategy. Firms that rarely turn away assets do not do that for sport. They do it when compliance teams decide the residual risk is no longer theoretical.
The manager most associated with the flagship product had already added disclosure that benefits could be disallowed after the fact. Nobody who read the documents can claim they were never told. Whether the live book already works around the concerns in the notice is unclear. The firm did not respond to reporters and has said in the past that it adapts strategies to stay inside guidance. That is the correct sentence for a lawyer to approve. It is not a map of the trades.
Would I put a client’s entire taxable book into a 250/150 long-short overlay whose main selling point is ordinary loss? Not on the current paper. Leverage between 130/30 and 250/150 is a market risk even if the tax theory holds. If the tax theory wobbles, you still have the market risk, plus an exit that may not be graceful.
The Companion Ruling On Fund Conversions
The same week brought a revenue ruling aimed at the other darling of the tax-alpha crowd: the Section 351 conversion into an exchange-traded fund. The pattern is easy to describe. An investor seeds a brand-new fund with a pile of highly appreciated stock. The fund then swaps that stock out through in-kind redemptions and ends up holding a diversified portfolio. If the form holds, the built-in gain never shows up on the investor’s return. More than 100 funds, with over $20 billion in seed assets, have launched this way since 2021, according to industry trackers. Crypto-oriented issuers have used related in-kind mechanics to sidestep income tests that would otherwise be awkward.
The ruling applies substance-over-form and step-transaction thinking. It treats the fund as a conduit and recharacterizes the sequence as a taxable exchange between the contributing investor and the authorized participant. That is a blunt instrument. It does not say every in-kind process in the market is tainted. It says a prearranged conversion that was never really a contribution to a genuine ongoing fund can be collapsed into the deal it actually was.
Form still matters. A sequence that has no business purpose except the tax result is the fact pattern these doctrines were written for.
Issuers and their counsel are now rereading seed documents. Some conversions will look like real launches with real follow-on assets. Others will look like a weekend bridge from a concentrated stock position to a diversified basket. The ruling does not need to catch every case to change the pipeline. It only needs to make the next seed conversation longer.
What Treasury Leadership Signaled In Public
As the notice went out, the Treasury secretary said publicly that the department is serious about cracking down on transactions designed to dodge taxes or exploit the federal code, and pointed to the companion ruling on prearranged fund contributions. For an administration that has cut taxes elsewhere, the move can look inconsistent at first glance. It is less odd if you separate rates from base. Cutting a rate and policing a shelter are different levers. A dollar of wage income offset by a currency forward is a dollar that does not arrive, and a dollar of borrowing that has to be sold to someone else while long yields sit near levels that already make budget math uncomfortable.
You can disagree with the priority. You can argue that legal tax planning is not dodging. The notice itself is more careful than the social-media line. It studies. It asks. It warns. Politics will supply the adjectives. Investors should supply the scenarios.
Three Paths The Market Is Already Pricing
This is a notice, not a regulation. That distinction will be repeated in every client letter between now and year-end. Behavior will not wait for the distinction to feel comforting. Three shifts are the ones I would watch.
- Product tweaks. Managers will lean on the stock-only long-short language that the notice treated more gently, and pull back on currency forwards and swap terminations that sit in the crosshairs.
- Slower inflows. Something on the order of a billion dollars a week had been moving into tax-aware long-short books. That pace is hard to defend while a retroactive question hangs over the K-1.
- Unwind risk. If clients head for the exit, leveraged books have to come down. Crowded shorts in names these quantitative portfolios favor are the place where a tax story becomes a market story.
The third point is the one equity desks should not ignore. A tax product with 200 percent gross exposure is still a portfolio. Deleveraging is not a press release. It is selling longs and covering shorts, sometimes in names that were never liquid enough for the crowd that piled in. I have watched “non-market” unwinds move prices before. The label does not protect the print.
What A Family Office Should Ask This Month
If you already own one of these funds, the useful questions are boring and specific. What fraction of last year’s loss was ordinary, and which instruments produced it? Is the manager still running currency forwards or swap terminations, or has that sleeve been shut? What does the offering document say about retroactive disallowance, and who bears the cost of a fight? How long would a full exit take at current borrowing and short-locate conditions?
If you were about to subscribe, the bar is higher than a backtest. A strategy that needs ordinary character to justify its fee has just been told that character is under review. Waiting through the comment window is not cowardice. It is sequencing. You can always add risk after the rules firm up. You cannot always unwind a partnership interest on the day a listing notice lands.
Questions worth writing down before the next capital call: 1. Source of ordinary character, instrument by instrument 2. Gross and net exposure, and the short book overlap 3. Exit timeline and any lock or gate 4. Disclosure on retroactive disallowance 5. Who pays if an audit expands
None of those questions require a view on politics. They require a view on your own liquidity. That is the part advisers sometimes skip when the slide says “tax alpha” in large type.
Listed Transactions And Why The Label Stings
A word on process, because the scare headlines skip it. Studying a strategy is not the same as listing it. A listed transaction carries disclosure duties, penalty exposure, and a chill that outlasts the final technical rule. A transaction of interest is a notch softer and still unpleasant. The notice says officials may designate. It does not say they have. Between those two sentences sits the entire comment period.
Firms will argue that long-short investing has a century of economic content, that currency hedges are real, and that swap terminations can be portfolio decisions rather than tax switches. Some of that will be true of some books. The trouble with an industrial product is that the facts start to rhyme. Same-day round trips. Terminations clustered in December. Loss ratios that dwarf the economic drawdown. When the pattern is the product, substance-over-form arguments get harder, not easier.
There is a reasonable counter. Congress writes character rules. Taxpayers are allowed to prefer the character the statute gives them. A currency forward really can produce ordinary loss under Section 988. A swap really can be a notional principal contract. The fight is not whether the code has shelves. The fight is whether stacking those shelves into a machine whose primary output is a wage deduction, with investment return as a byproduct, still counts as the activity the shelves were built for.
A Plain-Language Walk Through The Mismatch
Imagine two results arriving in the same year. Result A is a gain the code treats as capital. Result B is a loss the code treats as ordinary. If you can arrange to realize B against your salary and defer A, or realize A in a vehicle that does not flow the gain to you the same way, your tax bill falls even if your wealth did not. Scale that across thousands of accounts and you have an industry. The notice is aimed at the arranging, not at the accident of a bad stock year.
Foreign-currency forwards are a clean illustration because the character rule is old and specific. A contract that is a Section 988 transaction can throw off ordinary gain or loss. Do the round trip the same day, and the economic exposure may be tiny while the character is not. Selective swap terminations work differently. You close the contract that helps the loss line and keep the one that holds the winner. Identified straddles add a timing rule on top. None of this is new to derivatives counsel. What is new is the volume, the retail-adjacent distribution, and the explicit goal of shielding W-2 income.
Box spreads deserve their own sentence. A box can be built so the payoff resembles a Treasury bill. If the wrapper does not spit out current ordinary income, the holder gets a cash-like economic result with a different tax clock. Whether that survives contact with the notice is precisely the sort of question the comment file will be full of. I would not assume the answer is “yes” just because the payoff diagram looks elegant.
In-Kind Redemptions Are Not Automatically Suspect
It is worth saying the quiet half out loud. In-kind creation and redemption is how many exchange-traded funds avoid distributing capital gains to remaining holders. That mechanism is old, statutory, and used every day by plain index products. The notice and the ruling are not a surprise attack on the entire wrapper. They are aimed at uses that look pre-wired: seeding with appreciated stock solely to wash the gain, or leaning on redemptions to duck a qualifying-income test the fund could not otherwise meet.
The difference is intent and choreography, which is annoying because intent is what audits argue about. A fund that launches, gathers outside money, and rebalances through the normal authorized-participant pipe is in a different file from a fund that exists for a week as a tunnel between one holder and a diversified basket. Advisers who blur that line in client memos are not helping their clients.
Year-End Planning Just Got Less Cute
October is when tax conversations usually turn practical. Harvest this lot. Gift that lot. Fill the bracket. The notice lands in the middle of that season with a warning that some of the fancier tools may be rewritten later. That does not freeze ordinary planning. It should freeze the version of planning that only works if a derivative character rule is read in the most generous possible way.
A few habits still look sturdy. Owning investments long enough for long-term rates to matter. Harvesting genuine economic losses in stocks you would have sold anyway. Using retirement accounts for the income you do not want taxed now. Charitable gifts of appreciated shares when the gift was going to happen regardless. None of those require a swap termination calendar. They also will not produce a 28 percent ordinary loss on a portfolio that is up. That is the trade. The sturdy tools are smaller. They are also less likely to arrive with a retroactivity footnote.
If a manager tells you the notice “does not apply to us” without showing the instrument list, treat that as a start, not a finish. Specificity is the tell. Vague comfort is how these stories age badly.
The Deficit Math Behind The Timing
You do not need to like the framing to see the timing. Long yields near 5 percent make every sheltered dollar louder. Ordinary-income offsets reduce receipts today. Capital-gains deferral often reduces receipts later, and sometimes not at all if basis is stepped up at death. A Treasury that has to sell a lot of paper cares more about the today number than about a seminar on investment efficiency.
Critics will say enforcement energy aimed at legal structures is a substitute for harder fiscal choices. Supporters will say a code full of unintended machines is not a code. Both can be partly right. For the holder of the product, the debate is scenery. The exposure is the K-1, the leverage, and the exit.
How Managers Are Likely To Rewrite The Pitch
Expect the word “ordinary” to get quieter in sales decks and the phrase “long-standing stock technique” to get louder. Some funds will cap the derivative sleeve. Some will split share classes so the aggressive character sits in a vehicle only the most tax-motivated clients touch. Some will simply slow subscriptions until counsel finishes a memo. None of that is cynical. It is how a business survives a notice.
The harder rewrite is the one aimed at existing investors. If a 2025 loss was ordinary because of a forward book the manager no longer wants to run, does the 2026 product still do what the client bought? Fee language rarely promises a deduction. Marketing sometimes implied one. That gap is where complaints start, even before any auditor calls.
I would also watch the language in new subscription documents. Stronger risk factors are a gift, not an insult. A manager willing to write “benefits may be disallowed retroactively, and the fund may change instruments without notice” is closer to honest than a manager still leading with the holy-grail slide.
Crowded Shorts And The Market Side Of A Tax Story
Quantitative long-short books do not pick shorts at random. They cluster in factors: high momentum on the long side, weak quality or high short-interest names on the other, depending on the model. When many tax-aware funds run cousins of the same model, the short book becomes a shared exit. A redemption wave does not need a scandal to hurt. It needs a gate, a month-end, and a locate that disappears.
This is not a prediction of a squeeze. It is a reminder that tax products have market footprints. If inflows slow from a billion a week to something ordinary, the footprint shrinks slowly. If redemptions bunch, it shrinks rudely. Portfolio managers who are short the same names these books are short should know the overlap before they size up. That homework is dull. It is also how you avoid being the other side of someone else’s tax unwind.
What “Economic Substance” Will Mean In The Comment File
The phrase will be everywhere by November. Economic substance asks whether a transaction changed the taxpayer’s economic position in a meaningful way apart from tax, and whether the taxpayer had a substantial non-tax purpose. A hedged long-short book can answer yes on the first half. It has market exposure, borrowing cost, and the chance of a real drawdown. The second half is where funds that advertised the deduction first will struggle. Purpose is proved with memos, investment committees, and what the client was told. Decks that lead with the wage offset are exhibits, not decorations.
A fair reading leaves room for funds that harvest losses as a feature of a real return-seeking process. The notice almost says so when it nods at established stock techniques. The line it draws is fuzzy on purpose. Fuzzy lines are where conservative clients step back and aggressive ones look for the remaining gap. Both responses are rational. They are not the same risk budget.
A Note On Penalties Without The Horror Script
Nothing in this week’s paper assesses a penalty. Listing, if it comes, can bring disclosure penalties and accuracy penalties that sting. Retroactive guidance, if it comes, can reopen returns. Between here and there is a comment period, possible proposed regulations, and the usual litigation if someone with a large enough loss decides to fight. Most clients will never see a courtroom. They will see a manager letter, a smaller deduction, or a decision to exit.
That middle outcome is the one planning should assume until the paper says otherwise. Hope is not a tax position. Neither is panic. Document what you own. Ask what produced the character. Decide whether the after-tax story still works if next year’s loss is capital, or smaller, or delayed.
Rough sanity check: fee drag + financing + short rebate, then tax benefit only if character survives. If character is uncertain, price the benefit at a discount, not at the sales illustration.Who Actually Needed The Holy Grail
The client who already had large embedded gains could get a lot of relief from ordinary harvesting and from charitable or estate tools. The client who needed the new machine was the one with a big salary and a portfolio that had not compounded long enough to throw off gains. That is a huge population in a bull market that minted earners faster than it minted basis. It is also why the fiscal stake is larger than a hedge-fund niche suggests. Forty percent of a big firm’s assets is a clue. The addressable wage base behind those assets is the point.
There is a social argument hiding in the mechanics. A code that lets high earners manufacture wage offsets the median worker cannot access will draw political fire even when every form is filled in. You can believe the fire is unfair and still believe the fire is coming. The notice is the fire arriving in procedural clothing.
What I Would Tell A Client On A Tuesday Call
Keep the diversified taxable account. Keep harvesting real losses in names you are willing to exit. Do not add a leveraged ordinary-loss overlay until the comment period and the next piece of guidance have been read by someone who is paid to say no. If you are in one, get the instrument inventory and the exit math before you decide to stay. If your only reason for the allocation was the first-year deduction, that reason is now impaired, even if the portfolio return is fine.
I would also separate the fund-conversion question from the long-short question. They traveled together in the headlines. They are different facts. A prearranged seed of appreciated stock into a tunnel fund is on the wrong side of a ruling that already exists. A stock long-short book that never touched a currency forward may be on the side the notice itself described as familiar. Mixing them in one panicked email helps nobody.
And I would ignore anyone selling the next holy grail in the same week the last one was named in a notice. The pattern is old. The costume changes. The retroactivity sentence is what changed the costume’s price.
The Comment Window Is Short On Purpose
Late October is not a leisurely academic deadline. It sits before year-end allocations, before many partnership estimates, and before the stories clients tell themselves about January. Treasury knows that. So do the trade groups that will file. A thick comment file can slow a regulation. It rarely erases a warning that has already moved brokerage policy and subscription pace.
If you advise clients, the useful work this month is inventory, not prophecy. Which accounts used currency forwards. Which used swap close-outs. Which seeded a new fund with low-basis stock. Which merely run a long-short equity book with ordinary tax-lot management. Four piles. Four memos. One conversation that does not pretend they are the same trade.
A Longer View Than The Next K-1
Tax alpha, as a phrase, will survive. Investors should care about after-tax return. Turnover, location of assets, and lot selection are legitimate edges. The version that tried to industrialize ordinary loss against wages was always going to meet a skeptic with a badge. The surprise is not the skepticism. The surprise is how large the books got before the paper arrived, and how clearly some managers warned that the paper might reach backward.
When something seems too good to be true, the old line still works. In this case it may also be retroactively too good to be true. That is an uncomfortable sentence to put next to a return you already filed. It is a better sentence than discovering the discomfort in an audit letter two years from now.
Markets will adapt. They always do. Some capital will slide back toward plain direct indexing and toward funds that can explain their losses without a derivatives appendix. Some will wait for a final rule and then build the next structure one inch inside it. The inch is where the next notice will aim. For now, the practical edge is simpler than any model: know what you own, know what character it claimed, and do not let a sales illustration spend money the statute has not finished allocating.
If the guidance ultimately grandfathers old trades, the investors who paused will have paid a small opportunity cost. If it does not, the investors who treated the notice as noise will be explaining a deduction to someone who has already read the marketing deck. I know which conversation I would rather have.
